Targeting Designs of Scalable, Exploratory Summary Visualizations

Author: Sarikaya, Alper T (The University of Wisconsin – Madison)  2017,  Advisor: Gleicher, Michael L. 

Data visualization provides a human-digestible interface to digital data. With increasing data volumes and increased complexity and interrelationships, so have the demands on supporting effective visualization to support this interface. With complex data at scale, it becomes necessary to summarize the data in some manner to communicate high-level information of a dataset, such as distributions, trends, or anomalies. By the nature of summarization, fidelity in the visual representation of such summaries is reduced. With summarized data, visualization designs must make trade-offs to support particular types of tasks and analyses over others. In this dissertation, I present organizations and applications for the effective design of summary visualizations. Organizations of summary visualization identify the relevant factors that affect appropriate design, such as the method of summarizing data, the analysis goals and tasks of the viewer, and the characteristics of the data. Applications of summary visualizations demonstrate the holistic application of appropriate design decisions to support the analysis of complex scientific data. These factors are linked together to identify appropriate design strategies and highlight open problems for the effective design of visualization. Through this research presented herein, I provide new guidance for effective visualization of collections of data, allowing for the wider dissemination and analysis of complex data.